5854
J . O r g . Chem. 1994,59, 5854-5855
Chemoenzymatic Synthesis of 4-SubstitutedRiboses. S-(4'-Methyladenosyl)-~-homocysteine~ Carl R. Johnson,* John L. Esker, and Michael C. Van Zandt Department of Chemistry, Wayne State University, Detroit, Michigan 48202-3489 Received July 29, 1994@
Summary: The synthesis of 4-C-methyl-~-ribose,4-Cphenyl-D-ribose,D-ribose-4-d, and L-ribose derivatives as well as the title nucleoside by a chemoenzymatic strategy beginning from cyclopentadiene is described. The level of S-adenosyl-L-homocysteine (AdoHcy) (11, a natural feedback inhibitor of S-adenosyl-L-methioninedependent transmethylases, is controlled by AdoHcy hydrolase. The latter degrades AdoHcy by a NAD+dependent process involving (i) oxidation of the 3'hydroxyl, (ii) ,!?-eliminationof homocysteine, (iii) addition of water to the resulting enone, and (iv) re-reduction of the 3'-keto group to produce aden0sine.l
. -
H6
6H
1R=H 2R=CH3
Compounds which interfere with the transmethylation process have potential for therapeutic intervention in the viral replication process.2 Among possible mechanismbased inhibitors of interest to us is the AdoHcy analog 2 in which the H at the 4'-ribose position is replaced by a methyl group. Such a compound could not participate in the elimination step of the Abeles-Palmer pathway described above and thus may function directly or indirectly as an inhibitor of transmethylations. The implementation of this strategy required a convenient route to 4-substituted riboses either by manipulation of natural carbohydrates or by direct chemoenzymatic synthesis. As the former process presents a formidable task in the stereocontrolled introduction of a substitutent a t the configurational enter,^,^ we chose to develop a de novo chemoenzymatic strategy of 4-substituted riboses that would allow access to both the D and L series. We present here an efficient, stereoselective strategy to 4-substituted riboses involving the readily available enantiopure carbocyclic precursor 3 and its application to the synthesis of S-(4'-methyladenosyl)+homocysteine (2). +Dedicated to Prof. J. Bryan Jones on the occasion of his 60th birthday. Abstract published in Advance ACS Abstracts, September 1,1994. (1)Palmer, J. L.; Abeles, R. H. J.Biol. Chem. 1976, 251, 5817. (2) A direct correlation between antiviral potency of adenine analogs and their inhibition of AdoHcy hydrolase has been demonstrated (De Clercq, E.; Cools, M. Biochem. Biophys. Res. Commun. 1985,129,306. Keller, B. T.; Borchardt, R. T. Mol. Pharmacol. 1987, 31, 485). For recent reviews see: Wolfe, M. S.; Borchardt, R. T. J. Med. Chem. 1991, 34(51, 1521. Liu, S.; Wolfe, M. S.; Borchardt, R. T. Antiviral Res. 1992, 19, 247. (3)Yoshimura, J. Adv. Carbohydr. Chem. Biochem. 1984,42, 69133. Gough, B. M.; Gunner, S. W.; Overend, W. G. Carbohydr. Res. 1970. -. - -14. - , 173. (4) De Voss, J. J.; Hangeland, J. J.;Townsend, C. A. J. Org.Chem. 1994,59, 2715. @
7
Syntheses of enantiopure enone 3 have been developed in this and other laboratorie~.~ For the present purpose we prepared enone 3 from cyclopentadiene by a process involving the use of the very inexpensive crude porcine pancreatic lipase.6 The transformations of enone (+)-3, [a125D +66.3 (c 1.0, CHClS), into 4-methylribose is illustrated in Scheme l. a-Bromination of the enone was nearly quantitative. Excellent diastereofacial selectivity (>20:1 by NMR) was achieved in the 1,a-addition of methylmagnesium bromide to 4 a t -78 "Cto produce the alcohol 5. The ribose framework was unveiled from this carbocyclic precursor by ozonolysis of the vinyl bromide unit in methanollpyridine followed by reductive workup; 6 was isolated in 91% yield as a single anomer. The intermediate acid bromide was converted in situ to the methyl ester during ozonolysis. Refluxing 6 with MeOW acetonem+ gave the corresponding methyl glycosides (22: 1, ,!?:a),' and subsequent reduction of the methyl ester with lithium borohydride gave the protected ,!?-methyl 4-C-methylribofuranoside (7)8in 74% overall yield from 3. The condensation of 7 and N,N'-bis(trifluoroacety1)L-homocystine, dimethyl ester in the presence of trialkylphosphinesg proved unsuccessful, apparently due to the steric hindrance imparted by the 4-methyl group. The incorporation of the amino acid side chain was accomplished in good yield by formation of the 5-triflate and subsequent displacement with a suitably protected homocysteine in DMF. Initial studies utilized N,Ndiallylhomocysteine, methyl ester and N,N-dibenzy-homocysteine, methyl ester; however, neither the allyl nor benzyl amino protecting groups could be removed efficiently at the end of the synthesis. The coupling of N-(trifluoroacety1)-L-homocysteine,methyl esterlo with the 5-triflate was quantitative and proceeded without epimerization to yield 8. The acetonide was removed by ( 5 ) (a) Johnson, C. R.; Penning, T. D. J.Am. Chem. SOC.1988,110, 4726. (b) Hudlicky, T.; Natchus, M. G.; Nugent, T. C. Synth. Commun. 1992, 22, 151. (c) Siddiqi, S. M.; Schneller, S. W.; Ikeda, S.; Snoeck, R.; Andrei, G.; Balzarini, J.; De Clercq, E. Nucleosides Nucleotides 1993,12,185. (d) Ohrui, H.; Konno, M.; Meguro, H. Agric. Bwl. Chem. 1987,51,625. (e) Ali, S. M.; Ramesh, K.; Borchardt, R. T. Tetrahedron Lett. 1990,31, 1509. (0 Belanger, P.; Prasit, P. Tetrahedron Lett. 1988, 29, 5521. (g) Deardorff, D. R.; Shambayate, S.; Myles, D. C.; Heerding, D. J. Org. Chem. 1988,53, 3614. (6) Laumen, K.; Schneider, M. P. J. Chem. Soc., Chem. Commun. 1988, 1298. See also: Johnson, C. R.; Bis, S. J. Tetrahedron Lett. 1992, 33, 7287. (7) The minor anomer (15-20%) was observed when the reaction was run at 25 "C but was reduced to 3-5% under refluxing conditions. (8) The stereochemistry of the anomeric center in 7 was determined to be by the observed NOES between CH3-4 and H-l(O.9%)and the fact that no NOE was observed between H1 and H2 or H1 and H3. The anomeric center in 13 was determined to be a by the observed NOE's between H-5 and H-1 (0.4%)and between H-1 and H-2 (0.9%). (9)Nakagawa, J.; Hata, T. Tetrahedron Lett. 1976, 16, 1409. Serafinowski, P. Synthesis 1985, 926. (10) We found the preparation of N,N'-bis(trifluoroacety1)-L-homocystine dimethyl ester from L-homocystine by treatment with thionyl chloride in methanol followed by trifluoroacetic anhydride to be more efficient and convenient than that reported using successive treatments with dimethyl sulfite and trifluoroacetic anhydride (Cruickshank, P. A.; Sheehan, J. C. Anal. Chem. 1984, 36, 1191). Zindacetic acid cleavage of this disulfide gave the thiol, methyl N-(trifluoroacety1)-Lhomocysteine in 73%overall yield.
0022-326319411959-5854$04.50/0 0 1994 American Chemical Society
J. Org. Chem., Vol. 59, No. 20, 1994 5866
Communications
V"3
4
..~ d b
95%
-K-
5 6
7
YHBz
y
2
" ( a ) 0 2 , MeOH, rose bengal, thiourea hv; (b) AczO; (c) crude porcine pancreatic lipase. (d) benzyloxymethyl chloride; (e) OsOd, N-methylmorpholine N-oxide; (0 acetone, TsOH (g) H2, Pd-C; (h)pyridinium dichromate, 4 8, molecular sieves, CHzClz; (i) Brz, CHzClz 0 "C then Me& (1) MeOH, acetone, TsOH, 4 (m) LiBH4; (n) TfiO, Et3N, then Et3N; (i)MeMgBr, EtzO; (k) 0 3 , MeOH, pyridine, -78 -78 "C; (0)N-(trifluoroacety1)-L-homocysteine, methyl ester, NaH, DMF; (p) MeOH, HC1; (9)AczO, T E A (1)P-benzoyladenine, TMSOTf; (9) 0.2 M LiOH, 3 h; (t)N H 3 , MeOH, 12 h.
-
two successive treatments of 8 with acidic methanol, and glycosidation (1:20, P:a)' and reduction gave methyl the methyl 2,3-O-diacetylglycoside was prepared with 4-phenyl-a-~-ribofuranoside (13Yin a modest 20% yield from 3. acetic anhydride. Presilylated P-benzoyladenine and the above methyl glycoside were treated at elevated temperature (65 "C, CHsCN) with trimethylsilyl trifluoromethanesulfonate under Vorbriiggenl' conditions to give the fully protected P-nucleoside 9 in 64% yield from 8 (three steps, no a-anomer was isolated). Treatment of 9 with aqueous LiOH for 3 h followed by N H m e O H (to 10 11 12 13 efficiently remove the final N-benzoyl protection group) and purification by ion exchange chromatography using Compared to the manipulation of natural carbohyDowex 50 (H+) followed by Dowex 50 (NH4+)gave clean drates, this strategy offers the following advantages: (i) S-(4'-methyladenosyl)-~-homocysteine (2)in 98% yield the stereogenic center at C-4 is set prior to unveiling the (44% overall yield from 3). ribose framework; (ii) the availability of both enantiomers Various 4-substituted riboses should be accessible by of enone 3 allows access to either D or L riboses; and (iii) the strategy conveyed in Scheme 1. We found that a the method by which the ribose is unveiled and protected variety of nucleophiles including borohydrides, aryl Grigas the methyl glycoside circumvents problems associated nards, and alkyl Grignards provided excellent regio- and with pyranose-furanose equilibration, which for 4-subdiastereoselective additions to a-bromo enone 4. Synstituted riboses would undoubtedly lay in favor of the theses of ribosed derivatives are of particular interest, pyranose The biological activity of 2 will be for example, for the investigation of calicheamicin cleavreported in context with other mechanism based inhibiage of DNA by selective 5'- or 4'-hydrogen a b ~ t r a c t i o n . ~ tors based on the AdoHcy structure in a subsequent Treatment of 4 under Luche12 conditions with NaBD4 publication. Studies involving other 4'-alkylnucleosides provided 10. Subsequent ozonolysis, glycosidation (23: and nucleotides are in progress. 1,P:a), and reduction gave methyl 2,3-O-isopropylideneP-~-ribofuranoside-4-d (11)(95% deuterium incorporaAcknowledgment. This work was supported by a tion) in 67% overall yield from 3 (five-steps, not optimized). grant from the National Institutes of Health (CA37806) Similarly, methyl 2,3-O-isopropylidene-~-~-ribofuranoside and by a postdoctoral fellowship (J.L.E.) from Merck, was prepared from the enantiomer of 3.13 ChemoselecSharp & Dohme. tive ozonolysis of the vinyl bromide 12, prepared by Supplementary Material Available: Experimental proaddtion of phenylmagnesium bromide to 4, followed by cedures for key reactions and compound characterization data along with copies of the 'H a n d 13C NMR spectra for com(11) Vorbriiggen, H.; Krolikiewicz, IC;Bennua, B. Chem. Ber. 1981, 114, 1234 and references cited therein.
(12) Gemal, A. L.; Luche, J.-L. J.Am. Chem. Soc. 1981,103, 5454. (13) The enantiomer of 3 was prepared by manipulation of the cishydroxy acetate described in Scheme 1 or by a modification of the procedure by Siddiqi et a1.3cwherein an enzymatic resolutiodacylation of (f)-cis-4-phenoxy-2-cyclopenten-l-ol was carried out in isopropenyl acetate using Pseudomonas cepucia lipase. Methyl 2,3-O-isopropylidene-B-L-ribofuranosidewas prepared in 59% overall yield from the enantiomer of 3 when Luche reduction of 4 was carried out with NaB&. The optical rotation of the L-ribofuranoside was equal but of opposite sign to that of the well-known D-enantiOmer.
pounds 2, 7 , 8, 9, 11, a n d 13 (16 pages). This material is contained in libraries on microfiche, immediately follows this article in t h e microfilm version of the journal, a n d can be ordered from the ACS; see a n y current masthead page for ordering information.
(14)Angyal, S. J. Adv. Carbohydr. Chem. Biochem. 19@4,42, 15. from a L-erythroThe preparation of methyl 4-methyl-~-ribopyranoside pentopyranosid-4-ulosehas been reported (Overend, W. G.; White, A. C.; Williams, N. R. Carbohydr. Res. 1970,15, 185).
